SEP 14 1967
-/L~EFRZA R IE
AN EXPERIMENTAL INVESTIGATION OF ELECTRICAL
DISCHARGE MACHINING USING REVERSE POLARITY
by
Hans C. Juvkam-Wold
S.B. Massachusetts Institute of Technology
(1966)
Submitted in partial fulfillment
of the requirements for the
degree of Master of Science
at the
Massachusetts Institute of Technology
August, 1967
Signature of Author
,
Dept. of Mechanical Engineering, August 21, 1967
Certified by
Theis Supervisor
Accepted by
Chairman, Departmental Committee
on Graduate Students
AN EXPERIMENTAL INVESTIGATION OF ELECTRICAL
DISCHARGE MACHINING USING REVERSE POLARITY
by
Hans C. Juvkam-Wold
Submitted to the Department of Mechanical Engineering on
August 21, 1967 in partial fulfillment of the requirements
for the degree of Master of Science.
ABSTRACT
The principal purpose of this investigation was to determine
experimentally the parameters governing electrical discharge machining
(EDM) under conditions of reverse polarity (tool positive).
Tool steel
was machined on an Elox EDM machine using graphite and copper electrodes
(tools).
The frequency was varied from 0.34 kc to 63 kc, the duty cycle
from 30% to 80%, and the current from 1 to 115 amps.
The wear ratio (tool/work) was found to increase with frequency
and to decrease with duty cycle and with current.
It was found that
the wear ratio is correlated with the magnitude of the overcut, which
in turn is controlled by the size of the particles (usually hollow
spheres) eroded from the workpiece during each pulse of electrical
energy.
The metal erosion rate from the workpiece was found to be proportional to the 0.85th power of the current when graphite electrodes were
At a current of 35 amps, the maximum erosion rate of 0.022 cubic
used.
inches per minute was obtained at a frequency of 9 kc and 65% duty cycle.
The graphite tools consistently gave higher erosion rates and lower
wear ratios than those obtained using copper tools.
A qualitative theory of the erosion mechanism is suggested and
examined in the light of the experimental data.
Thesis Supervisor:
Title:
Robert E. Stickney
Associate Professor of Mechanical Engineering
-2-
ACKNOWLEDGEMENTS
I wish to thank Professor R. E. Stickney for suggesting this
research topic and for his guidance and encouragement throughout.
I
also wish to thank Mr. T. Viswanathan for making the runs with copper
electrodes, and Mrs. Rose Hurvitz for her expert typing of the manuscript.
My understanding of the EDM process has been greatly enhanced
through discussions with my fellow students:
and Viswanathan.
De Nigris, Doret, Weetman,
These discussions have been invaluable.
Further I wish to thank the Elox Corporation of Michigan for
supplying the EDM machine and power supply, and for giving advice and
helpful suggestions on a number of occasions.
I also wish to express my deep appreciation to Mene Grande Oil
Company (Gulf) for four and a half years of continuous fellowship
support.
Finally, I wish to thank the Research Laboratory of Electronics
for supplying me with an RLE fellowship during this last summer.
-3-
TABLE OF CONTENTS
Page
Abstract
2
Acknowledgements
3
List of Figures
6
1.
INTRODUCTION
8
2.
EXPERIMENTAL APPARATUS AND PROCEDURE
3.
4.
10
2.1.
Electrical Discharge Machine
10
2.2.
Power Supply
11
2.3.
Tool and Workpiece Description
11
2.4.
Measurement of Tool Wear and Workpiece Erosion
12
2.5.
Determination of Gap Voltage, Frequency, and Duty Cycle
12
2.6.
Current Measurement
13
EXPERIMENTAL PROGRAM:
OBJECTIVES AND RESULTS
14
3.1.
Outline of Experimental Program
14
3.2.
Experimental Results
15
18
DISCUSSION OF RESULTS
4.1.
Theoretical Background
18
4.2.
Conditions Determining Wear Ratio
20
4.21.
Frequency
20
4.22.
Duty Cycle
21
4.23.
Current
22
4.24
Tool Material
23
4.25.
Overcut
23
4.26.
Size of Particles Eroded from the
Cathode in Each Pulse
-4-
23
-1
Page
4.3.
4.4.
Conditions Determining Erosion Rate from the Workpiece
24
4.31.
Frequency
24
4.32.
Duty Cycle
24
4.33.
Current
25
4.34.
Tool Material
25
Photomicrographs
26
5. SUMMARY
6. RECOMMENDATIONS
REFERENCES
APPENDIX A:
Data Tables
APPENDIX B:
Graphs and Photomicrographs
-5-
LIST OF FIGURES
Page
Fig. 1. Schematic view of experimental apparatus under reverse
10
polarity conditions.
Fig. 2. Geometry of tool and workpiece.
11
Fig. 3. Variation of voltage across the electrode gap when the
unit is not machining, as observed on the oscilloscope.
Fig. 4. The anode spot size increases with gap spacing.
13
18
Fig. 5. Picture on the oscilloscope under good machining
20
conditions.
Fig. 6. Dependence of wear ratio on frequency, at a current of
35 amps and a duty cycle of 50%.
37
Fig. 7. Dependence of wear ratio on duty cycle, at a frequency
of 9 kc and a current of 35 amps.
Fig. 8. Dependence of wear ratio
38
on current, at a frequency of
9 kc and a duty cycle of 50%.
39
Fig. 9. Dependence of wear ratio on overcut at a current of 35
amps (all frequencies and duty cycles).
Fig.10.
40
Dependence of overcut on the volume of metal eroded
from the workpiece in each pulse (for all frequencies,
duty cycles, current, and tool materials).
Fig.l1.
41
Dependence of wear ratio on the amount of metal eroded
from the workpiece in each pulse (for all frequencies,
duty cycles, and currents).
Fig.12.
42
Dependence of erosion rate on frequency, at a current
of 35 amps and a duty cycle of 50%.
-6-
43
Page
Fig. 13.
Fig. 14.
Dependence of erosion rate on duty cycle, at a
frequency of 9 kc and a current of 35 amps.
44
Dependence of erosion rate on machining current,
at a frequency of 9 kc and a duty cycle of 50%.
45
(10
in
-4
Fig. 15.
Dependence of erosion rate
3
on current,
at a frequency of 9 kc and 50% duty cycle.
Fig. 16.
Erosion rate does not depend on the depth of cut in
the region tested.
Fig. 17.
47
Photomicrograph (100x) showing copper tool with
steel deposit from workpiece (1 kc).
Fig. 18.
48
Photomicrograph (100x), showing copper tool with
erosion craters (68 kc).
Fig. 19.
46
48
Photomicrograph (500x), showing graphite tool
with protective deposit of steel from workpiece
(0.34 kc).
Fig. 20.
48
Photomicrograph (200x), showing graphite tool with
slight steel deposit (63 kc).
Fig. 21.
48
Photomicrograph (100x), showing graphite tool
covered with splashes of steel and hollow steel
spheres (0.34 kc).
Fig. 22.
49
Photomicrograph (100x), showing steel spheres from
EDM fluid.
This cross-sectional view clearly
shows that at least some of the spheres are hollow.
-7-
49
1. INTRODUCTION
The erosive effect of electric sparks has been known for over 200
years.
During the last quarter century a machining process making use
of this effect has been developing.
This process, known as electrical
discharge machining (EDM) consists essentially of passing successive
pulses of electrical energy between two electrodes, the tool and the
workpiece, and gradually eroding the latter to any desired configuration.
A concise description of the process is given by Cook
extensive evaluation by Berghausen et al.2
and an
EDM has many advantages
over conventional chip machining, and is particularly useful where intricate shapes or very hard materials are to be machined.
Since the tool
need never be in physical contact with the workpiece, it can be made
from some easy-to-shape soft material like graphite, which incidentally
has outstanding wear characteristics, and any conducting material can
be machined, irrespective of hardness.
In spite of the importance of EDM today, the basic principles
underlying the process are far from being fully understood.
Until this
handicap is overcome, the process cannot be developed to its full potential.
A systematic approach to this problem is now well under way here
in the Mechanical Engineering Department at M.I.T., and this thesis is
the second of a series.
As indicated above, tool wear is a matter of great importance,
since tool replacement is a major operating expense.
Under standard
polarity EDM, the workpiece is the positive electrode (anode), but
experience has recently shown that under certain operating conditions,
tool wear can be substantially reduced, and sometimes almost completely
eliminated, by reversing the polarity of the electrodes, thereby making
the tool be the anode.
The purpose of the present investigation is to
determine empirically the parameters governing machining performance
of EDM under conditions of reverse polarity.
Particular emphasis is
placed on determining the wear ratio (the volume of material worn away
from the tool per unit volume of material eroded from the workpiece),
-8-
a
and the machining or erosion rate of the workpiece, under different operating conditions.
By observing the conditions under which different tool
wear is obtained, it should be possible to understand better the wear
mechanism, and in turn to extend the region of low tool wear.
In Section 2 of this thesis the apparatus is described.
Also the
experimental procedure is described in some detail.
The experimental results are presented in Section 3, and each graph
is briefly described.
An attempt has also been made to explain the results,
but this is left until Section 4. The independent variables are the frequency of the electric pulses,the duty cycle (the ratio of the time that
the power is on to the total period of each pulse, expressed as a percentage), the average current, and the tool material.
The most important
dependent variables are wear ratio and erosion rate (in this report the
latter expression always refers to the workpiece).
Section 4 contains a general discussion of the results, and a provisional theory is presented 4n an attempt at explaining the results.
The
limitations of this theory should be emphasized, and may not be valid under
conditions different from the ones tested here.
-9-
2.
EXPERIMENTAL APPARATUS AND PROCEDURE
The apparatus consists essentially of an EDM machine, a power
supply, miscellaneous measuring devices, and an oscilloscope.
A
schematic of the setup is shown in Fig. 1.
E DM
Fig. 1.
2.1.
F I u id
Schematic view of experimental apparatus
under reverse polarity conditions.
Electrical Discharge Machine
The EDM machine used in these experiments is an Elox HRP-63 unit.
This is a rigid and accurate machine.
The workpiece is located inside
a workpan, under at least three inches of EDM fluid (ofter called the
"dielectric"; in this case Eloxol #13 was used throughout).
The tool is
attached to a ram whose vertical motion is controlled by a servomechanism which in turn is controlled by the electrode gap (the distance between
the tool and the workpiece).
The machine also has a pump and filter for
-10-
providing a constant supply of clean EDM fluid to the electrode gap.
Another pump is provided for quick filling of the workpan.
2.2.
Power Supply
Square voltage pulses are supplied from an Elox SSD 400 NW
experimental power supply.
This unit can supply pulses at frequencies
from 0.34 kc to 170 kc, and currents up to 400 amps.
The duty cycle
(percent ON-time) can be varied continuously from about 8% to 90%. The
power supply also controls the feed rate of the servomechanism, and
permits some variation of the electrode gap.
2.3.
Tool and Workpiece Description
The geometry of the tool and workpiece is shown in Fig. 2. The
tool is a cylindrical rod of 0.75-inch diameter.
The workpiece is
Too I
Workpiece
0.75"
dia m.
1.5''x 1.5" x 0.5"
EDM
Fluid
Fig. 2. Geometry of tool and workpiece
-11-
Flow
1.5"x 1.5"x 0.5", and has a 3/16" hole drilled at the center to permit
the flow of coolant (EDM fluid).
This geometry was chosen because of
the cylindrical symmetry which permits good flushing and simple analysis.
The material of the workpiece is tool and die steel, H-13, produced
by the Carpenter Steel Company under the number 883.
Both hardened and
unhardened steels were tested, with surprisingly little difference in
performance.
copper;
Three kinds of tool material were tested:
(1) 99.9% pure
(2) Kost Kutter No. 8; and (3) Kost Kutter No. 12.
The last
two types are commercial graphite EDM electrodes produced by the Speer
Carbon Company.
KK #8 is a standard electrode material, and KK #12 is
somewhat harder and stronger, and is specifically recommended by the
producers where fine detail on the electrode is desired.
2.4.
Measurement of Tool Wear and Workpiece Erosion
The linear wear on the tool and on the workpiece were measured at
four different points around the circumference of a circle of radius 1/4"
and axially centered.
was used.
An Ames No. 384 dial gauge, correct to 0.0005",
As very little rounding off of the edges of the electrodes
occurred, the volume of material removed was obtained by multiplying
the linear wear by the cross-sectional area (0.415 in2 ).
The overcut was taken to be half the difference between the diameter
of the hole eroded in the workpiece and the final diameter of the tool.
For these measurements a Helios inside-outside vernier caliper, accurate
to
0.001", was used.
Again an average of four readings was taken. Some
of these readings were later checked using a screw-type Fowler micrometer.
2.5.
Determination of Gap Voltage, Frequency, and Duty Cycle
The variation of voltage (V) across the electrode gap with time (t)
was observed on a Tektronix Type 511A Cathode Ray Oscilloscope (Fig. 3).
-12-
I---
t I.
* -----
t2
Fig. 3. Variation of voltage across the electrode gap when the
unit is not machining, as observed on the oscilloscope.
The frequency of the pulses is simply the reciprocal of the total time
per pulse
frequency = t
+
1
.
The duty cycle, which is defined to
2
be the percentage ON-time, can also be obtained from the oscilloscope
%
tl
I!duty cycle = z x 100% .
2
The oscilloscope sweep speed was calibrated
'
using a Tektronix Type 181 Time-Mark Generator, accurate to 0.01%.
2.6.
Current Measurement
The machining current was measured with a Weston D.C. ammeter con-
nected across an external shunt.
This ammeter was calibrated against a
Hewlett Packard Model 412A DC Vacuum Tube Voltmeter, and also against a
Tektronix Type 536 Oscilloscope.
off by more than 1.5%.
The readings on the ammeter were never
The currents mentioned in this report are taken
directly from the ammeter.
-13-
-. a
3. EXPERIMENTAL PROGRAM:
3.1.
OBJECTIVES AND RESULTS
Outline of Experimental Program
The overall purpose of this program was to enhance our understanding
of the EDM process, in an attempt at improving the performance of the process.
It was felt that this could best be achieved through an empirical
investigation of the parameters which influence wear ratio and machining
rates.
The "standard" test was chosen to be a frequency of 9 kc, a duty
cycle of 50%, and a current of 35 amps.
Hence, when the dependence of
wear ratio and erosion rate on frequency was to be determined, the duty
cycle was kept constant at 50%, and the current at 35 amps.
Similarly,
the dependence on duty cycle was investigated at 9 kc and 35 amps, and
so on.
9 kc was chosen because near this frequency the maximum erosion
rate of material from the cathode is obtained.
In addition, sufficient
wear is obtained on the tool to allow fairly accurate measurement of this
wear.
50% duty cycle was chosen because this value was obtainable from
the power supply over a wide range of frequencies, from 0.34 kc to 34 kc.
Barring this, a higher duty cycle would have been preferred, but 77% was
obtainable only up to 9 kc.
A current of 35 amps was arbitrarily chosen,
and was easily realizable over most of the spectrum tested.
However, a
lower current, about 20 amps, would have been obtainable over an even
larger range of conditions.
Graphite is probably the most popular EDM tool material in use
today, but considerable amount of work has also been done with copper.
For this reason two commercial graphite EDM electrodes, and also pure
copper, were chosen for these experiments.
Another reason for choosing
copper is that visual observation of the erosion craters and material
deposits is much easier on copper than on graphite.
Any material which will conduct electricity can be machined by EDM,
but since it was impossible in the time available to test all the materials
that fall into this category, the choice was narrowed down to a hard substance that is often machined by EDM, and tool and die steel, H-13, was
-14-
chosen.
3.2.
*
Both hardened and unhardened H-13 were tested.
Experimental Results
The wear ratio (tool/work) increases with increasing frequency, and
decreases with increasing duty cycle or increasing current.
Figure 6 shows the dependence of wear ratio on frequency.
For a
current of 35 amps and a duty cycle of 50%, and for frequencies between
0.34 kc and 63 kc, the following results were obtained:
tool, the wear ratic
with a graphite
is proportional to the 0.8th power of the frequency;
and with a copper tool, the wear ratio is proportional to the 0.14th
power of the frequency.
That is,
graphite tool: (wear ratio) % (frequency)0.8
copper tool:
(wear ratio)
(1)
'I,(frequency) 0.14
(2)
Over the whole range investigated, graphite gives a lower wear ratio
than copper, although if the trends indicated on the graph (Fig. 6) continue
to much higher frequencies, then copper would give a lower wear
ratio than graphite at frequencies above 250 kc.
The relationship between wear ratio and duty cycle is shown in Fig. 7.
At a frequency of 9 kc and a current of 35 amps, the following results were
obtained:
With a graphite tool, for duty cycles between 30 and 80%, the
wear ratio decreases as the log of the duty cycle increases:
graphite tool:
(duty cycle) = 100 exp[- 14.4(wear ratio)]
(3)
With a copper tool under the same conditions, the wear ratio increases
slightly with duty cycle between 30% and 50%, but for duty cycles between
50% and 80%, the wear ratio decreases with increasing duty cycle.
Figure 8 shows that at a frequency of 9 kc and a duty cycle of 50%,
the wear ratio decreases with increasing current, for currents between 1
and 100 amps.
This holds both for graphite and copper tools.
This rela-
tionship is again logarithmic and may be expressed as follows:
*
Currently in progress is a study to investigate the effect of
different workpiece materials by T. Viswanathan, research assistant, M.I.T.
-15-
graphite tool:
(current) = 140 exp[-29(wear ratio)]
(4)
The wear ratio decreases exponentially with increasing overcut
[-(hole diam. - tool diam.)] (Fig. 9).
At a current of 35 amps, for
frequencies between 0.34 kc and 63 kc, and for duty cycles between 30%
and 80%, the following relationships hold approximately:
graphite tool:
copper tool:
(wear ratio) = 0.6 exp[-690(overcut in inches)]
(wear ratio) = 0.8 exp[-160(overcut in inches)]
(5)
(6)
As shown in Fig. 10, the overcut is proportional to the cube root
of the volume of cathode material eroded per pulse.
For all conditions tested:
(overcut) = 5.8(volume eroded)1/ 3
where overcut is measured in inches, and volume eroded is measured in
in 3/pulse.
Note that this relationship holds for all frequencies between 0.34
and 63 kc, duty cycles between 30 and 80%, currents between 1 and 100 amps,
and for both graphite and copper tools, i.e., it holds under all conditions
investigated in this program!
As the volume of material eroded from the workpiece per pulse becomes
larger, the wear ratio decreases, and may even go negative.
shown in Fig. 11.
This fact is
The data points on this graph do not follow any simple
equation.
The erosion rate of material from the workpiece increases with
increasing frequency, goes through a maximum, and then decreases again
(Fig. 12).
These experiments were carried out at a current of 35 amps and
a duty cycle of 50%.
For copper electrodes the optimum frequencv appears
to be about 1-8 kc, and the erosion rate decreases quite significantly at
frequencies below 1 kc or above 20 kc.
With graphite electrodes the maxi-
mum erosion rate was obtained between 4 and 14 kc, and highly reduced rates
were obtained for frequencies below 2 kc or above 20 kc.
Most of the experiments discussed in this report were carried out
using unhardened steel H-13 as workpiece material.
However, a number of
runs weremade using KK #8 tool and hardened steel H-13 workpiece.
-16-
Under
(7)
most experimental conditions slightly lower erosion rates and slightly
lower wear ratios were obtained with the hardened H-13.
However, at
frequencies below 4 kc, significantly lower erosion rates are obtained,
as shown in Fig. 12.
Near 9 kc the difference is very small, and as all
the other tests in the program were carried out at this frequency, the
data points with hardened and unhardened steel can be mixed without
increasing the scatter significantly, except in Fig. 13.
More details
regarding this matter can be found in the data tables in Appendix A,
Tables 1-a and 1-b.
In Fig. 13 the dependence of workpiece erosion rate on duty cycle
at 9 kc and 50% duty cycle is shown.
There is little variation, except
for a reduced rate at very low or (in the case of hardened steel workpiece) at very high duty cycles.
At a frequency of 9 kc and a duty cycle of 50%, the erosion rate
increases with current over the whole range tested (Fig. 14).
For cur-
rents between 7 and 70 amps the following equations hold:
graphite tool:
copper tool:
(8)
(erosion rate) ' (current) 0.85
(erosion rate) % (current)
(9)
Currents above 70 amps could not be obtained with graphite tools.
With
copper as tool material, a substantial reduction in the rate of increase
of erosion rate with current is obtained above 70 amps.
Another way of presenting the erosion data is often used, so in
3
Fig. 15 the erosion rate, in units of (\ 10
amp.min)
current.
is plotted against
The erosion "efficiency",as it might be called, is seen to
decrease with increasing current in the case of graphite tools.
With
copper as tool material, this erosion "efficiency" is essentially constant
up to about 60 amps, after which it falls off significantly.
In order to check the dependence of the erosion rate on the depth of
cut, a couple of runs were made.
The results speak for themselves (Fig. 16).
For a depth of cut less than 0.4" the erosion rate is substantially constant.
-17-
4. DISCUSSION OF RESULTS
Theoretical Background
4.1.
In order to explain fully the experimental results obtained in this
investigation, it is essential to give some theoretical background.
attempt is made here to survey all the theories related to EDM.
No
Instead
a selection of theories which tend to support our experimental results is
given.
We here suggest that the wear of the tool (anode) is determined by
the size of electrode gap.
One reason for this is that the stream of
electrons that leaves the cathode, gradually spreads out due to electron
repulsion and electron diffusion, as shown in Fig. 4. This is by no means
a novel idea, and has been suggested by Ullmant;, 3 Webb,4 and others.
Hence, as the gap spacing is increased, the area of the anode spot also
increases, so that for a given total current the power density becomes
less.
When the power density is low enough, the ON-time is not suffi-
ciently long to raise the temperature of the tool surface up to the melting
point, and no wear results by this mechanism.
TOOL
(+)
A node
Gap
Spacing
Cathode
Spot
WORKPI E CE
(-)
Fig. 4. The anode spot size increases with gap spacing.
-18-
Mandel'shtam and Raiskii5 suggest that the anode surface is eroded
by "vapor jets" emitted from the cathode surface, and the cathode is
similarly eroded by smaller jets from the anode.
They find that by
increasing the electrode gap the wear on the anode decreases, and at
large enough spacings the anode becomes plated with material from the
cathode.
This method of wear from the anode appears to be correct, but
the erosion from the cathode is probably predominantly caused by positive
ion bombardment.
"...
Kesaev6 sums it up as follows:
Thus, it may be seen that the cause of electrode erosion is
the predominantly thermal and mechanical effect exerted on the
electrodes by the charged (ions] and neutral (vapor jets] particles
moving at high velocity in the discharge gap."
Combining these ideas, we suggest that the anode wear mechanism
consists primarily of vapor jets emitted from the cathode, a mechanism
whose efficiency decreases with increasing gap distance;
whereas the
cathode is eroded primarily by positive ions which are accelerated in
the cathode fall (a potential drop of the order of 10 volts, very close
to the cathode surface), a mechanism in essence independent of electrode
spacing.
The energy balance at the cathode is thus of great importance
for determination of the erosion rate at the cathode (workpiece), but
this subject has been discussed in detail by Holm,7 Somerville,8 Kesaev, 9
and others, so it will not be repeated here.
To summarize the above, when the electrode spacing increases, the
anode wear definitely decreases, whereas the cathode wear remains essentially constant.
Before going on to see how the above theories can explain the
experimental results, we shall define some terms that will be used later.
To do this, let us take a look at the picture on the oscilloscope (Fig. 5)
when the machine is operating.
break down the gap.
down the gap.
across the gap.
Line 1 shows voltage pulses that do not
Line 2 shows a broad band of pulses that do break
The height of line 2 represents the instantaneous voltage
This voltage, V 2 , multiplied by the instantaneous current
(peak current), Imax, gives the instantaneous power during the pulse.
total energy per pulse is V 2 1maxt 1
-19-
The
where t1 is the ON-time of the pulse.
1
1
2
V2
3
--
0
0
N
t I
F F
t2
Fig. 5. Picture on the oscilloscope under good machining conditions.
Of this energy, some goes to the cathode, some to the anode, and some to
the EDM fluid.
area
The instantaneous power reaching the anode, divided by the
of the anode spot (Fig. 4), gives the instantaneous power density at
the anode surface.
D.C. ammeter.
The average current is the current that is read on the
This current multiplied by V2 gives approximately the
average power put out by the power supply.
4.2.
4.21.
Conditions Determining Wear Ratio
Frequency
As the frequency is lowered, each spark erodes more material from the
cathode (workpiece) but less material from the anode.
-20-
At first this may
seem contradictory, but these results can be explained.
When larger par-
ticles are removed from the cathode (because of higher energies per pulse),
the overcut becomes larger because these particles have to be washed out
of the hole being drilled.
If the overcut is not sufficiently large, then
breakdown will occur between the side of the -tool and the side of the hole
in the workpiece, through the particle.
These same large particles will
similarly cause the electrode gap to be larger, probably of the same order
of magnitude as the overcut, and definitely proportional to the overcut.
But when the electrode gap is larger, the power density at the anode
becomes less since the arc has a chance to spread out considerably due
to electron repulsion.
Hence, the anode surface is heated up less, and
may not reach the melting point of the tool material.
The above theories claim that the wear on the anode is caused by
"vapor jets" of material ejected from the cathode at high velocities.
At
low spacings (high frequencies) the vapor jets will wear down the anode,
whereas at larger spacings (lower frequencies) the molten metal tends to
plate onto the anode rather than wearing it down.
This is why we get
such low wear ratios at low frequencies (Figs. 6, 17, and 19), because
the steel from the cathode now protects the tool surface.
At higher fre-
quencies, and hence lower spacings, very little protection is offered
(Figs. 6, 18, and 20).
De Nigris10 has calculated the time required to
wear away such a protective layer at an assumed incident power density
6
2
of 10 watts/cm .
From the above discussion it should be clear why the wear ratio
increases with frequency.
4.22.
Duty Cycle
The wear ratio decreases with increasing duty cycle.
time it is not quite clear why this should be so.
constant, so the power input is constant.
At the present
The current is kept
This gives a constant energy
per pulse, so the size of the particles removed from the cathode, and
hence the gap distance, should be constant.
-21-
This should, in turn, give a
constant wear ratio.
One possibility is that a long pulse might remove
more material than a shorter pulse of equal energy, but then the erosion
rate should increase consistently with duty cycle, and this is not so,
as Fig. 13 shows. Another possibility is that the machinability (the
fraction of the total number of pulses put out by the power supply that
actually do break down the gap) decreases with duty cycle.
Then, in an
effort to maintain a constant average current, the peak current during
the pulses which do break down the gap becomes higher, so more energy
is supplied for each effective pulse.
The particles removed are now
larger, and hence the wear ratio lower, thereby explaining the shape of
the curve.
Yet another possibility is suggested by the picture on the
oscilloscope (Fig. 5).
Line 2 sometimes extends across the OFF-time as
well, and apparently this could be due to the arc not being extinguished
after each pulse.
Then, for these pulses twice as much energy (or more)
goes into each one, so much larger particles are removed, thereby giving
large spacings and low wear ratios.
As the duty cycle increases, the
OFF-time gets shorter and shorter, so it seems reasonable to assume that
more and more of the arcs would fail to extinguish, thus bringing about
the decrease in wear ratio that Fig. 7 shows.
However, the last sugges-
tions are purely speculative and must be supported or denied by further
experimentation.
One shred of evidence is given by the fact that, in
general, the overcut appears to increase with duty cycle (Appendix A,
Table 1-a).
This is not positively shown by the data, however, and will
not be until a more accurate way of measuring the overcut is developed.
4.23.
Current
For currents up to 70 amps the wear ratio decreases with increasing
current, both for graphite and for copper tools (Fig. 8).
Qualitatively
similar results were obtained by Livshits et al.11 but the conditions of
their experiments were not fully specified, so a direct comparison is
impossible.
The reason for the relationship shown in Fig. 8 is that at higher
-22-
currents the energy per pulse is higher.
This causes larger particles
to be removed from the workpiece, so the gap spacing is larger, and
hence the tool wear and wear ratio are lower.
4.24.
Tool Material
As mentioned earlier, graphite has considerably better wear charac-
teristics than copper, over the whole range of conditions included in
Figs. 6 through 8. Graphite KK #12 gives a very slightly lower wear
ratio than KK #8 but this difference is almost insignificant.
At 50% duty cycle, copper has a wear ratio of more than 0.2,
even at the lowest frequencies tested (0.34 kc).
By increasing the
duty cycle to 80% and keeping the current at 15 amps, essentially no
wear is obtained on the copper tool at frequencies below 1 kc (tests
made by Elox Corporation of Michigan).
4.25.
Overcut
Above it was assumed that the wear ratio is determined by the
overcut.
Fig. 9 substantiates this assumption.
At a constant current
of 35 amps, but irrespective of frequency and duty cycle, wear ratio
is seen to decrease with increasing overcut.
With KK #8 as tool
material, less than 1% wear ratio is obtained whenever the overcut is
greater than 5 x 10-3 inches.
4.26.
Size of Particles Eroded from the Cathode in Each Pulse
It was also assumed above that the overcut is governed by the size
of the particles eroded from the workpiece in each pulse.
ports this assumption.
Fig. 10 sup-
One might then expect the overcut to be propor-
tional to the diameter of these particles.
If this is true, then the
overcut should be proportional to the cube root of the volume of each
3
particle, since the particles are spherical and volume "' (diameter)
-23-
In calculating the volume of metal eroded in each pulse, 100% machinability was assumed throughout.
Also, there is no reason why the metal
eroded in each pulse should not come out as a number of separate particles
rather than as one separate particle.
tionship is remarkable!
In view of these facts the rela-
The good agreement shown in Fig. 10 is probably
due to the fact that the energy per pulse puts an upper bound on the size
of the particles eroded, and the overcut is probably determined by the
largest of these particles rather than by their average size.
This cor-
relation (Fig. 10) holds under all conditions tested!
From the above discussion one should expect a close correlation
between the wear ratio and the volume of metal eroded per pulse.
Fig. 11
bears out this expectation.
4.3.
4.31.
Conditions Determining Erosion Rate from the Workpiece
Frequency
A fairly flat peak in the erosion rate versus frequency curve (Fig.12)
is obtained between 2 and 14 kc, at 35 amps and 50% duty cycle.
For fre-
quencies below 1 kc the erosion rate is seen to drop off substantially. It
is believed that the main reason for this is unstable machining conditions.
At frequencies above 20 kc the erosion rate drops off even more rapidly. At
these high frequencies the ON-time for each pulse is quite short, and as
discussed by De Nigris,10 it takes a finite time to bring the machined surface up to melting temperatures.
This is probably the main reason for the
reduced erosion rate here.
4.32.
Duty Cycle
At 9 kc and 35 amps there is no substantial variation of the erosion
rate with duty cycle (Fig. 13).
A slight reduction at very high and very
low duty cycles can probably be attributed to unstable machining.
Parti-
cularly low erosion rates are obtained at high duty cycles when hardened
tool steel is machined (Fig. 13).
The reason for this is not quite clear.
-24-
4.33.
Current
The erosion rate, measured in units of in 3/min, increases with
increasing current (Fig. 14).
This is not surprising, since increasing
the current means increasing the power input.
However, the erosion rate
in units of in 3/(amp.min) decreases slightly with increasing current, when
graphite electrodes are used (Fig. 15).
With graphite as tool material
it was impossible to achieve stable machining at currents above 70 amps.
With copper electrodes currents up to 115 amps were obtained, but a
substantial drop in the erosion "efficiency" (Fig. 15) occurs at about
60-70 amps.
The principal cause for this is believed to be copper plating
of the steel workpiece.
Because of these high currents, the power density
at the anode is now sufficient to create the anode vapor jets mentioned by
Mandel'shtam and Raiskii,5 but these jets do not substantially erode the
cathode surface.
Due to the large spacing they tend to plate onto the
cathode, thereby reducing the effectiveness of the other cathode erosion
mechanism (positive ion bombardment).
This brings about the reduced
erosion "efficiency" at high currents.
4.34.
Tool Material
Figs. 12 through 15 consistently show that graphite tools give
higher erosion rates under all conditions tested (except for the case of
hardened steel (Fig. 12) which is not applicable in this comparison since
no runs were made with copper tools on hardened steel).
Of the graphite
tool materials, the KK #12 gives slightly higher erosion rates than KK #8,
but this difference is only of the order of a few percent.
The main
advantage of the KK #12 appears to be its greater strength and resistance
to chipping, particularly in cases where fine detail on the tool is
required.
However, it is true that both
slightly lower wear ratios and
slightly higher erosion rates are obtained with this tool material.
-25-
4.4.
Photomicrographs
At this point it appears profitable to take a close look at the
photomicrographs in Figs. 17 through 22.
theories presented.
These pictures support the
Under so-called "NO WEAR" conditions, which means
low frequencies and high duty cycles, with small or zero wear ratio, the
gap spacing between the tool and workpiece is always large.
Under these
conditions the "vapor jets" emitted from the cathode are ineffective in
wearing down the tool.
In fact, they tend to deposit, or plate, onto the
tool and protect it from wear.
Fig. 17 shows a blob of cathode metal on
a copper tool, and Fig. 19 shows large splashes of cathode material on a
graphite tool.
The wear mechanism now has to remove this protective layer
before the anode material can be worn away.
This plating action is so
effective that sometimes a negative wear ratio is obtained, due to a
plating rate which is higher than the wear rate.
The buildup on the
tool is usually more pronounced in places where many sparks hit - in
regions where the tool is usually worn most - near and at corners and
edges.
picture.
In Fig. 5 two distinct voltage lines are shown on the oscilloscope
Apparently one of these lines (No. 2) represents the potential
drop across the gap when the arc is between steel on the cathode and
copper (or graphite) on the tool, and the other line (No. 3) shows the
potential when the arc is active between steel on the cathode and steel
on the tool surface.
This appears to be a reasonable deduction, but will
be investigated further.
At higher frequencies tool wear can be considerable.
Now the gap
spacing is lower and the cathode jets are slowed down less before they
strike the surface of the tool.
This high-energy molten or vaporized
metal will usually not deposit on the tool but will erode a crater there
instead.
The result is tool wear.
Figure 18 shows such craters on a
copper tool and in Fig. 20 some craters and a small steel deposit are
shown on a graphite tool.
A large portion of the material eroded from the workpiece is carried
away in the form of hollow spheres by the EDM fluid.
Some of these spheres
may also be found deposited on the tool (Fig. 21) or on the workpiece.
-26-
To get a closer look at these metal spheres, some of them were embedded
in plastic and then polished. The result gave a cross-sectional view
of the spheres, which indeed may be hollow (Fig. 22).
The fact that these spheres do exist, tends to support Zolotylch's 1 2
theory that the metal is primarily removed in molten form, rather than
in the form of vapor (separate molecules).
Zolotylch claims that most
of the metal is removed from the workpiece after the arc is extinguished.
This is not an impossible theory, since during the arc the pressure outside the surface reaches many atmospheres - a convincing argument for
this is given by Holm
-
and when the arc is shut off, the pressure in
the arc suddenly drops to about 1 atmosphere.
With respect to this new
pressure, all the metal in the molten pool at the cathode surface finds
itself superheated, and explodes violently.
This may be the mechanism
for creation of the "vapor jets" which wear down the anode, and is probably the source of the hollow spheres.
However, more studies are
required before this can definitely be pinned down as the mechanism of
erosion.
-27-
5. SUMMARY
This investigation shows that graphite is a better tool material
than copper for machining tool and die steel by EDM under conditions
of reverse polarity.
The wear ratio (tool/work) is found to increase
with increasing frequency, and to decrease with increasing current or
duty cycle.
These results can all be explained through the concept of
electrode spacing.
At small spacings the vapor jets from the cathode
erode the tool and bring about high wear ratios.
At large spacings
these same vapor jets tend to plate and protect the tool surface,
giving rise to low wear ratios.
The magnitude of the gap distance is
controlled by the diameter of the largest particles eroded from the
cathode, and this is in turn governed by the energy per pulse.
Hence
the wear ratio is essentially determined by the energy per pulse.
The erosion rate reaches a maximum value at frequencies from 1
to 14 kc, depending on the tool material.
The duty cycle has little
effect on the erosion rate; when the average current and the frequency
are kept constant, a small peak is obtained at about 65% duty cycle.
Erosion rate increases with current, as one might suspect, but for
currents above 70 amps, with copper tools, substantial plating of the
cathode surface occurs, thereby reducing the erosion rate.
It should be noted that a low frequency alone is not sufficient
to guarantee a low wear ratio.
If the current or the duty cycle is
low enough, substantial tool wear will occur even at the lowest frequencies obtainable with today's EDM machines.
Low wear ratios may
be obtainable at higher frequencies if the gap spacing can be kept
large enough.
-28-
6. RECOMMENDATIONS
This study shows that in order to obtain low wear ratios, it is
essential to have large gap spacings.
At the present time this is achieved
mainly through the removal of large particles from the workpiece, at the
expense of poor surface finish.
Research should be done to find another
way of maintaining large gap spacings, possibly through higher breakdown
voltages, or EDM fluids with lower breakdown strengths, or both.
Another
possibility for maintaining large spacings might be to introduce artificially into the EDM fluid large conducting particles, so as to eliminate the need for removing large particles from the workpiece.
As a first step towards the above suggestions, a chamber will be
constructed for the purpose of collecting and measuring all the particles
formed during amachining test. Other variables to be investigated in
the near future include the pressure, temperature, consistency, and volume
flow rate of the EDM fluid.
The effect of different workpiece materials
will also be studied (currently in progress), to see if the theories suggested here are applicable over a wider range of conditions.
The effect
of duty cycle variation is still not fully understood, and will be further
investigated, possibly by means of measuring the machinability.
This may
be done by comparing the average current (from the ammeter) with the peak
current (from the scope);
and the results can be checked by making use of
an electronic counter to measure the actual number of pulses over a fixed
period of time.
Surface damage to the workpiece has always been of great importance
for some applications.
layer.
Microcracks are often formed in the resolidified
This problem would be avoided if all the molten metal could be
removed, rather than resolidified on the surface.
It may be possible to
achieve this through variations in the shape of the peak current profile.
However, before this is attempted, a better model for the metal removal
mechanism should be developed.
-29-
REFERENCES
1. Cook, N. H., Manufacturing Analysis, Addison-Wesley Publishing Co.,
Inc., Reading, Massachusetts, 1965, pp. 121-126.
2. Berghausen, P. E., Brettschneider, H. D., and Davis, M. F.,
"Electro-Discharge Machining Program," Technical Documentary Report
No. ASD-TDR-7-545 of the Cincinnati Milling Machine Co., Ohio, 1963.
3. Ullmann, W., "Electro-Erosive Metalworking-Theory and Practical
Application of a Proved Method,"
Technische Rundschau No. 8, Feb. 28,
1958.
4. Webb, R. S., "High Voltage - Reverse Polarity EDM," Patent No.
3,158,728, filed July 26, 1960.
5. Mandel-shtam, S. L. and Raiskii, S. M., "Mechanism of Electroerosion of Metals," Izvestiya Akad. Nauk SSSR, Ser. Fiz., Vol. 13,
1949, No. 5, pp. 549-565.
6. Kesaev, I. G., Cathode Processes in the Mercury Arc, Consultants Bureau,
New York, 1964 (Russian Original, 1961), p. 29.
7. Holm, R., Electric Contacts Handbook, Springer-Verlag/Gottingen/Heidelberg,
1958, pp. 467-469.
8.
Somerville, J. M., The Electric Arc, Barnes and Noble, Inc., New York,
1959, p. 66.
9. Kesaev, I. G., Cathode Processes in the Mercury Arc, Consultants Bureau,
New York, 1964 (Russian Original, 1961), pp. 59-61.
10.
De Nigris, E. G., "Heat Transfer Aspects of EDM," S.M. Thesis, M.I.T.,
August, 1967 (in preparation).
11.
Livshits, A. L., Bannikov, G. K., and Sigarev, A. M., "High-Resistance
Electrodes for Electro-Pulse Processing of Metals," Stanki-I-Instrument,
Vol. 29, No. 5, 1958, pp. 23-25.
12.
Zolotylch, B. N., "The Mechanism of Electro-Erosion of Metals in a Liquid
Dielectric Medium," Problemy Elek. Obrabotki Materialov, 1960.
13.
Holm, R., Electric Contacts Handbook, Springer-Verlag/Gottingen/Heidelberg,
1958, pp. 462-463.
-30-
APPENDIX A
DATA TABLES
-31-
TABLE 1-a:
Run No.
Frequency
kc
Tests using graphite Kost Kutter No. 8 tool and hardened steel H-13 workpiece.
Duty
Cycle
Current
amps
0.41
0.81
9.14
16.7
63.2
0.34
0.68
1.40
3.80
8.5
17.0
33.6
6.8
7.6
9.2
9.0
9.0
9.0
9.0
8.4
9.0
9.6
9.7
9.5
35
35
35
35
11
23
23
35
35
35
35
23
1.2
35
35
11.5
69
23
58
1.2
46
35
35
35
-0-
Wear Ratio
3
in /in
3
-0.0067
-0.0026
0.0063
0.0696
0.228
0.0078
0.0029
0.0005
0.0141
0.0418
0.0598
0.117
0.319
0.0782
0.0316
0.0912
0.0133
0.0692
0.0205
0.16
0.029
0.0645
0.0311
0.0179
Erosion
Rate
-3 3
10 in
min
6.8
8.05
12.0
14.3
1.20
5.48
6.83
11.7
16.2
17.1
16.1
9.0
0.39
14.3
19.7
7.12
28.5
13.0
25.3
0.207
22.1
18.0
19.0
16.2
Erosion
Rate
10 -4 in 3
amp.min
1.94
2.01
3.43
4.08
1.14
2.38
2.97
3.34
4.63
4.88
4.02
3.91
3.25
4.08
5.63
6.18
4.13
5.65
4.37
1.72
4.81
5.15
5.43
4.63
Volume
Eroded
-8 3
10 in
pulse
28.6
16.5
2.21
1.42
0.032
27.2
16.2
14.0
7.08
3.36
1.57
0.447
0.096
3.13
3.56
1.32
5.27
2.40
4.67
0.041
4.10
3.12
3.26
2.85
Overcut
10 -3in
9.5
8.5
6.0
3.1
1.3
8.5
7.3
6.0
5.5
3.2
3.0
1.5
1.5
2.5
4.5
3.8
3.5
3.0
4.0
1.8
3.8
3.0
4.0
5.0
TABLE 1-b:
Run No.
Frequency
ke
B- 1
B- 2
B- 4
B- 5
B- 6
B- 8
B- 9
B-10
B-11
B-12
B-13
B-14
B-16
B-1 7
B-19
B-20
B-21
B-22
B-2 3
B-24
B-25
B-26
B-27
B-29
B-31
B-32
0.34
0.69
1.4
3.8
9.6
30
17
60
8.6
9.0
9.3
9.4
8.9
9.3
9.8
9.3
8.6
8.6
8.7
8.7
8.9
8.9
9.2
9.2
9.2
9.1
Tests using graphite Kost Kutter No. 8 tool and unhardened steel H-13 workpiece.
Duty
Cycle
%
A?/
Current
Wear Ratio
amps
in /in3
28
35
35
35
35
23
35
11.5
35
35
35
35
35
35
35
35
11.5
17
23
29
35
46
58
1.2
69
7
0.00625
0.00385
0.00276
0.0241
0.0554
0.123
0.0696
0.179
0.0827
0.0765
0.0533
0.0557
0.0338
0.0237
0.0178
0.0466
0.0828
0.0756
0.0712
0.0690
0.0496
0.0404
0.0214
0.161
0.0205
0.0942
Erosion
Rate
Erosion
Rate
Volume
Eroded
10 -3.in 3
min
10 -4 in 3
amp.min
10 -8in 3
pulse
10- 3in
10.2
13.5
15.0
17.2
17.4
10.6
17.9
2.20
15.0
21.0
16.7
20.5
21.5
20.6
18.1
21.7
7.60
9.95
17.4
18.6
25.0
26.2
0.189
32.5
3.64
3.86
4.28
4.92
4.97
4.61
5.11
1.91
4.28
6.0
4.77
5.85
6.15
5.89
5.17
6.20
6.60
5.85
6.00
6.00
4.65
5.43
4.52
1.57
4.72
4.56
6.52
49.6
32.4
18.2
7.58
3.02
0.54
1.77
0.056
2.91
3.88
2.95
3.62
4.03
3.68
3.08
3.89
1.47
1.93
2.63
3.33
3.47
4.65
4.66
0.034
5.85
0.833
8.5
7.5
8.2
4.8
2.8
2.0
2.7
1.2
2.7
3.7
4.3
3.5
4.8
4.3
4.6
3.9
3.9
3.3
3.4
3.5
3.4
3.5
3.8
2.2
4.2
3.5
13.8
110,
RIO
Overcut
TABLE 2:
Run No.
Frequency
kc
201
202
203
204
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
0.33
0.68
1.4
3.9
17
32
63
8.3
8.3
8.2
8.6
8.6
8.7
9.0
8.7
8.6
8.6
7.5
7.5
7.9
8.0
8.1
8.1
8.0
8.1
Test using graphite Kost Kutter No.
Duty
Cycle
Current
amps
35
35
35
35
35
23
1.2
35
35
35
35
35
35
35
35
35
35
7
9.2
11.5
17
23
29
35
46
12 tool and unhardened steel H-13 workpiece.
Wear Ratio
in3/in
3
0.0055
0.0
0.00076
0.0240
0.0547
0.104
0.324
0.0728
0.0618
0.0667
0.0467
0.0440
0.0437
0.0407
0.0325
0.0173
0.0143
0.0851
0.0838
0.0814
0.0744
0.0666
0.0618
0.0412
0.0361
MI
Erosion
Rate
Erosion
Rate
Volume
Eroded
10-3 3
min
104 in3
amp.min
10-8 3
pulse
11.85
14.25
16.35
19.2
14.1
10.4
0.461
14.2
15.6
20.7
16.9
17.9
21.2
21.4
22.2
21.0
20.7
4.72
5.93
7.24
10.7
14.9
18.7
17.9
27.1
3.39
4.07
4.67
5.48
4.03
4.52
3.84
4.06
4.46
5.92
4.83
5.12
6.05
6.12
6.34
6.00
5.92
6.75
6.45
6.29
6.30
6.48
6.45
5.12
5.90
59.0
34.8
20.1
8.14
1.39
0.542
0.0122
2.84
3.11
4.21
3.28
3.50
4.05
3.94
4.25
4.07
4.03
1.05
1.32
1.52
2.23
3.06
3.85
3.74
5.57
Overcut
in
10 -3.
10.3
9.0
7.5
4.3
2.8
2.2
1.5
3.5
3.5
4.2
3.7
3.5
4.7
4.5
5.0
6.1
6.5
4.0
4.2
4.5
4.2
4.3
4.8
4.1
5.0
TABLE 3:
Run No.
Frequency
Duty
Cycle
kc
52
53
54
55
56
57
58
62
65
66
67
68
69
71
72
73
74
75
76
78
79
80
81
85
86
87
89
90
91
93
95
96
98
0.34
0.68
1.3
3.9
9.3
17
33
17
9.3
9.9
8.8
8.7
9.8
9.3
8.2
8.7
8.7
8.7
8.7
8.8
8.9
8.9
8.5
9.1
9.0
9.0
8.5
8.3
17
0.67
8.4
7.3
7.0
50
50
50
50
50
50
50
50
60
78
50
40
70
70
35
45
30
50
50
50
50
50
30
50
50
50
40
35
50
50
50
30
50
Tests using copper tool and unhardened steel H-13 workpiece.
Current
Wear Ratio
amps
in 3/in3
38
36
36
38
36
36
18.4
36
36
36
35
36
36
36
35
35
36
70
93
23
115
11.5
35
11.5
58
0.254
0.252
0.324
0.440
0.514
0.391
0.468
0.320
0.438
0.350
0.504
0.498
0.373
0.376
0.467
0.492
0.473
0.450
0.450
0.517
0.432
0.527
0.587
0.540
0.362
0.562
0.487
0.489
0.391
0.297
0.532
0.626
0.371
17.5
36
35
35
35
35
35
47
Erosion
Rate
-3 3
10 in
min
9.6
11.6
13.7
13.5
12.9
12.8
5.9
12.1
13.8
14.1
12.5
13.5
13.5
14.4
14.8
12.0
13.9
amp.min
2.5
3.3
3.9
3.8
3.6
3.6
3.2
3.4
3.9
4.0
3.6
3.8
3.7
4.1
4.2
3.4
3.9
3.2
2.5
4.2
2.1
3.2
3.5
3.5
4.0
3.2
4.1
4.1
4.0
4.9
3.7
2.7
4.2
22.2
23.5
9.6
24.6
3.7
12.3
4.0
23.1
5.9
14.5
14.4
13.8
17.1
13.0
9.6
19.6
M -1,1,
Erosion
Rate
10 -4 in 3
1!1, Oil
10
Volume
Eroded
-8 3
10 in
pulse
Overcut
47.6
28.2
17.3
5.72
2.32
1.29
0.30
1.21
2.48
2.38
2.37
2.58
2.29
2.58
3.00
2.30
2.85
4.25
4.51
1.83
4.58
0.69
2.43
0.73
4.27
7.0
7.0
7.2
4.8
2.5
3.8
2.5
4.5
3.8
4.5
3.0
3.0
4.8
3.0
3.5
3.2
2.5
2.25
2.5
3.0
3.0
3.8
3.5
4.5
4.5
4.5
4.0
4.4
3.8
9.3
4.0
3.8
6.5
1.09
2.95
2.91
1.37
42.7
2.58
2.20
5.64
1"
10-3 in
APPENDIX B
GRAPHS AND PHOTOMICROGRAPHS
-36-
1.0
0-
0
101
0
4-
0
10-2
l13L_
10~
Fig. 6.
1.0
10
FREQUENCY
102
( kc )
Dependence of wear ratio on frequency at a current of 35 amps and a duty cycle of 50%.
0
0.4
0
0
4-
0.3
0.2
0.1
0.0
20
30
DUTY
Fig. 7.
40
50
60
70
80
CYCLE
Dependence of wear ratio on duty cycle at a frequency of 9 kc and a current of 35 amps.
90
100
0.6
0
0.4
0
0
6--
0.2
0.0
1.0
10
CURRENT
Fig. 8.
100
(amps)
Dependence of wear ratio on current at a frequency of 9 kc and a duty cycle of 50%.
1.0
~
__________
0__
0
_
_
_
_
_
_
_
_
_
0
ono
S]00
100
OA
.
A
10-
00
LU
1102
MATERIAL
TOOL
A
Graphite
KK*8
o
Graphite
KK#12
o
Copper
0
2
4
OVERCUT
Fig. 9.
8
6
(10~3
10
in)
Dependence of wear ratio on overcut at a current of
35 amps (all frequencies and duty cycles).
-40-
I
I
10~ 2
io-
3
10-6
10-8
V OLUME
Fig. 10.
E RODE D
(in3
pul se)
Dependence of overcut on the volume of metal eroded from the workpiece in each pulse
(for all frequencies, duty cycles, currents, and tool materials).
0.8
0
0
0
ne
0.4
0
0.2
LU
0.0
-0.2
10-10
10~9
VOL UM E
Fig. 11.
108
E RODED
io
7
10-6
( in3/pulse )
Dependence of wear ratio on the volume of metal eroded from the workpiece in each pulse
(for all frequencies, duty cycles, and currents).
TOOL
MATERIAL
Graphite
KK
Graphite
KK #12
#
8
WORKPIECE
MATERIAL
Steel
un-hardened
H-13,
C
E
3
O
I
Copper
A
_
Graphite
KK
# 8
hardened
,
0.02
0.01
|_||L
0.00
0.2
0.5
W II_
1.0
F R EQUE NCY
Fig. 12.
0-
20
100
50
( kc )
Dependence of erosion rate on frequency at a current of 35 amps and a duty cycle of 50%.
[Some of these runs were made at lower currents and then corrected to 35 amps using Eq. (8).1
0 100"'POW-0111"0011",
I
It
Ithl I- I WA
It I Ili
-i-
Copper
0
EG
LU
i
r aphi t e
KK #8
-
0.02
-
,
hardened
m
so,
o
0
0
0.01
0
LU
30
40
DUTY
Fig. 13.
60
50
CYCLE
70
(%)
Dependence of erosion rate on duty cycle at a frequency of 9 kc and a current of 35 amps.
80
4
C
E
C
0.02
0.01
0.005
5
10
20
CURRENT
Fig. 14.
50
100
200
( am ps )
Dependence of erosion rate on machining current at a frequency of 9 kc and a duty cycle of 50%.
10
... CEL
oE
-
o
CURRE NT
Fig. 15.
Dependence of erosion rate
(
104 in
amp.min -
(amps)
o
on current at a frequency of 9 kc and 50% duty cycle.
0.5
0.4
0
-
C
RUN
#
15
RUN
#
16
0.3
U
00.2-U-
~0.1
I-
0
2
TIME
Fig. 16.
10
8
6
4
(min
12
14
16
)
Erosion rate does not depend on the depth of cut in the region tested.
W
W
o
I
I
:
L
"_
Photomicrograph (100x),
Fig. 17.
showing copper tool with steel deposit
Note that no serious
from workpiece.
damage to tool occurred (low frequency
conditions, 1 kc).
Fig. 18. Photomicrograph (100x),
showing copper tool with erosion
craters (high frequency conditions,
68 kc).
Fig. 19. Photomicrograph (500x),
showing graphite tool with large
protective splashes of steel from
workpiece (low frequency conditions,
0.34 kc).
Fig. 20. Photomicrograph (200x),
showing graphite tool with slight
steel deposit, offering very little
protection to tool (high frequency
conditions, 63 kc).
-48-
A
Fig. 21. Photomicrograph (100x),
showing graphite tool covered with
splashes of steel and hollow steel
spheres (low frequency conditions,
0.34 kc).
-49-
Fig. 22. Photomicrograph (100x),
showing steel spheres from EDM fluid.
This cross-sectional view clearly shows
that at least some of the spheres are
hollow.